Age-related changes in muscle strength are an important contributor to frailty in the elderly. The time course and causes for these changes are poorly understood. This project examines the changes across the adult life span with an emphasis on what occurs early in middle age when the changes begin. The research consists of three parts. First is an ongoing study in the BLSA comparing concentric (shortening) and eccentric (elongation) phases of movement assessed at slow, fast and zero (isometric) speeds.Almost 500 male and female subjects from the 20s through the 80s have been tested and longitudinal data collection is beginning. Age associated declines are found in cross-sectional analysis for both concentric and eccentric testing for all muscle groups and speeds. The cross-sectionally determined rate of decline is greater in the quadriceps femoris than for the biceps brachii. The rates of decline in concentric and eccentric strength tend to be parallel in men. Concentric strength declined to a much greater extent than eccentric strength in women. Second is an analysis of twenty-five years of muscle and power measurements in the BLSA. The initial analysis examines the issue of how qualitative changes in muscle mass are related to strength. Creatinine excretion declined in a manner similar to muscle strength, while body mass index, cross sectional area and lean body mass showed less change.The study suggests that measures that examine muscle contractile components may show a stronger association with strength than cross sectional muscle area. Third are studies to examine the peripheral nerve contribution to strength loss. A protocol has been developed that explores motor unit function at different levels of muscle exertion in the quadriceps femoris. The technique shows reasonable test-retest reliability with a .1-.3 coefficient of variation. The size of recruited motor units increases at increasing percentage of maximal voluntary contraction. The study will examine the effects of the motor unit size and firing pattern.